mcp: improve http transports error handling and make transport work with any size message

[alexbumbacea]

We have a use case for messages larger than 1MB, hence making this configurable would help.
In order to discover the source of our problems we had to spend some time investigating, as errors were silently discarded. So we also included changes to improving surfacing errors .

[findleyr]

Let's discuss this change in #726. I'm not sure we shouldn't just remove the buffering.

[alexbumbacea]

@findleyr current proposal (no fixed buffer size) approach is inline with most sdks. Can we get this merged?

[findleyr]

@alexbumbacea sorry for the delay, will review tomorrow.

[alexbumbacea]

@findleyr all comments have been either replied to or addressed. are there any other concerns?

[alexbumbacea]

@findleyr is there anything else?

[alexbumbacea]

@findleyr can we get this merged?

[findleyr]

@alexbumbacea very sorry, I thought I had merged this. Total oversight on my part.

@maciej-kisiel we may want to cherry-pick this into the release, or do a second prerelease.

[findleyr]

Apologies again for this getting lost in the weeds: I think there are a few things that need to be addressed before merging.

[findleyr]

It looks like my comments were addressed. However could you please also address @maciej-kisiel's before we merge?

Thanks for this change, this is a good cleanup.

[alexbumbacea]

@findleyr @maciej-kisiel can we get this merged and included in a prerelease? Thanks!

[maciej-kisiel]

It looks good to me, I wanted to give @findleyr some time to take a look since there was a small change from the version he approved. But I think we can merge it and adjust as a follow-up if needed.

With regard to the release, I think I would prefer to only include fixes to bugs that were introduced with a given pre-release. 1.3.0-pre.1 is already soaking some time and I wouldn't like to prolong the release more than necessary. This change will then be included in 1.4.0 which is planned ~in the last week of February.